摘要
针对传统的模糊C-均值聚类算法对初始聚类中心较敏感、易陷入局部最优的缺点,将粒子群优化算法和FCM算法相结合,提出一种改进的模糊聚类算法。该算法利用粒子群算法的全局搜索能力代替FCM算法寻找初始聚类中心,使其跳出局部最优,实现模糊聚类。主要从反映数据集分类的类内紧致性程度和类间分离性程度的角度考虑,重新设计适应度函数。实验结果表明,提出的算法在聚类正确率和有效性指标上有更好的效果。
Aiming at the problem of traditional fuzzy C-means clustering algorithm that it is sensitive to the initial clustering centers and easy to fall into the local optimization, an improved algorithm that combines Particle Swarm Optimization algorithm with FCM algorithm is proposed. Depending on utilizing the global searching ability of Particle Swarm Optimization algorithm instead of the FCM algorithm, the new algorithm searches the initial cluster centers and escapes from the local optimization so as to achieve fuzzy clustering at last. Meanwhile, it mainly redesigns the fitness function from the perspective of compactness in intra-class and separation in inter-class. The experimental results show that the proposed algorithm has a better effect on both the cluster validity indexes and clustering accuracy.
出处
《计算机工程与应用》
CSCD
2013年第22期115-118,122,共5页
Computer Engineering and Applications
基金
国家自然科学基金(No.61103129
No.61202312)
江苏省科技支撑计划资助项目(No.BE2009009)
